Activated Protein C Induces Endothelial Cell Proliferation by Mitogen-Activated Protein Kinase Activation In Vitro and Angiogenesis In Vivo
Activated protein C (APC), a natural anticoagulant, has recently been demonstrated to activate the mitogen-activated protein kinase (MAPK) pathway in endothelial cells in vitro. Because the MAPK pathway is implicated in endothelial cell proliferation, it is possible that APC induces endothelial cell proliferation, thereby causing angiogenesis. We examined this possibility in the present study. APC activated the MAPK pathway, increased DNA synthesis, and induced proliferation in cultured human umbilical vein endothelial cells dependent on its serine protease activity. Antibody against the endothelial protein C receptor (EPCR) inhibited these events. Early activation of the MAPK pathway was inhibited by an antibody against protease-activated receptor-1, whereas neither late and complete activation of the MAPK pathway nor endothelial cell proliferation were inhibited by this antibody. APC activated endothelial nitric oxide synthase (eNOS) via phosphatidylinositol 3-kinase–dependent phosphorylation, followed by activation of protein kinase G, suggesting that APC bound to EPCR might activate the endothelial MAPK pathway by a mechanism similar to that of VEGF. APC induced morphogenetic changes resembling tube-like structures of endothelial cells, whereas DIP-APC did not. When applied topically to the mouse cornea, APC clearly induced angiogenesis in wild-type mice, but not in eNOS knockout mice. These in vitro events induced by APC might at least partly explain the angiogenic activity in vivo. This angiogenic activity of APC might contribute to maintain proper microcirculation in addition to its antithrombotic activity.
- activated protein C
- endothelial protein C receptor
- protease-activated receptor-1
- mitogen-activated protein kinase
Activated protein C (APC) is a serine protease that plays a central role in physiological anticoagulation. Activation of protein C occurs on the endothelial cell surface by thrombin bound to thrombomodulin.1 The endothelial protein C receptor (EPCR), which is a glycoprotein on the vascular endothelium, binds to protein C, thereby augmenting protein C activation by the thrombin-thrombomodulin complex.2 APC regulates the coagulation cascade by inactivating activated forms of factors V and VIII in the presence of protein S,3 and it exerts a profibrinolytic effect by inactivating plasminogen activator inhibitor-1.4 The importance of APC as a natural anticoagulant is illustrated by the development of severe thrombosis in patients with congenital protein C deficiency.5
In addition to the anticoagulant effect, APC has been shown to be capable of activating protease activated receptor (PAR)-1 on endothelial cells,6 thereby activating the mitogen activated protein kinase (MAPK) pathway. The MAPK pathway is activated during mitosis, meiosis, and G0-G1 transition7 and is implicated in regulation of the cell cycle, mitogen-induced cell growth, and proliferation.8 The MAPK pathway plays an important role in signal transduction from growth factor receptors to the nucleus, thereby regulating cell proliferation. These observations strongly suggest that APC might induce endothelial cell proliferation via MAPK activation. Furthermore, because endothelial cell proliferation is a prerequisite step in angiogenesis, it is possible that APC induces angiogenesis.
In the present study, we examined these possibilities using cultured human umbilical endothelial cells (HUVECs) and the mouse cornea angiogenesis assay. Results strongly suggested that APC can induce angiogenesis, probably by activating the MAPK pathway in endothelial cells. Thus, it is likely that APC is a multifunctional protease that might be critically involved in regulation of microcirculation not only by anticoagulation, but by the promotion of neovascularization.
Materials and Methods
Protein C was purified from human plasma, activated by thrombin, and purified by cation exchange chromatography as described previously.9,10 Purified APC displayed a single band on sodium dodecylsulfate polyacrylamide gel electrophoresis.10 Thrombin activity was not detected in the purified APC concentrate. APC was inactivated by using diisopropyl fluorophosphate (DIP) as described previously.9 Lω-Nitro-l-arginine methyl ester (l-NAME) and wortmannin were purchased from Sigma. KT5823, a selective inhibitor of cyclic GMP-dependent protein kinase (PKG), and LY-294,002 were purchased from Alexis corporation. Heat-inactivated supplemented calf serum (SCS) from HyClone Laboratories. Anti-ERK1/2 antibody, anti-phosphorylated ERK1/2 antibody, anti-MEK1/2 antibody, and anti-phosphorylated MEK1/2 antibody were purchased from Cell Signaling. Monoclonal anti-PAR-1 antibody (ATAP2) was obtained from Santa Cruz Biotechnology. All other reagents were of analytical grade.
HUVECs were isolated from fresh human umbilical cord veins as described previously.11 Cells were grown to confluence at 37°C in humidified 95% air/5% CO2 as described previously.11 Cells were used for experiments up to passage number 4.
5-Bromodeoxyuridine Incorporation Assay
HUVECs were seeded in 96-well plates at 80% confluence (1×104 cells/well) and incubated in M199 with 10% SCS for 24 hours. APC and various pharmacological agents were then added, and after 16 hours, 5-Bromodeoxyuridine (BrdU; 10 μg/mL) was added to each well. Two hours later, the culture media were removed, and the BrdU incorporation was estimated using a BrdU incorporation assay kit (Hoffman La Rocsh) according to manufacturer’s instructions.
Endothelial Cell Proliferation Assays
Endothelial cell proliferation was monitored by a technique described by Hood et al12 with some modification. Briefly, HUVECs (5×103/well) were incubated with APC and various pharmacological agents in M199 containing 10% SCS in 96-well plates. Although culture media containing 0.2% serum were used in the original assay method, HUVECs in culture media containing 1% serum were demonstrated to be apoptotic at 18 hours after incubation.13 Thus, we used culture media containing 10% SCS to avoid apoptosis in the present study. After a 4-day incubation, cell number was estimated using a MTT kit (Hoffman La Rocsh) according to manufacturer’s instructions.
Matrigel Morphogenetic Assay
Formation of endothelial tube-like structures was analyzed using growth factor reduced MATRIGEL matrix (Becton Dickinson). Wells of 48-well culture cluster dishes were coated with 100 μL/well and allowed to solidify for 30 minutes at 37°C. HUVECs, after being starved for 24 hours in M199 with 10% SCS, were seeded on polymerized Matrigel (2×104 cells/well) and further propagated for 16 hours, then fixed in PBS containing 3% formaldehyde and 2% sucrose. To quantify the length of tube-like structures, three random phase-contrast photomicrographs (3 objective) per well were taken, and the length of tube-like structures was measured. Tube-like structures were quantified as the average total length of tube-like structures per low-power field.
Measurement of MEK1/2, Phosphorylated MEK1/2, ERK1/2, Phosphorylated ERK1/2, eNOS, and Phosphorylated eNOS
HUVECs were cultured in 6-well dish. On the day of experiments, culture media was changed M199 with 1% SCS containing various test drugs. After 30-minute incubation period, APC was added to culture media. After indicated time, whole cell lysates were collected and subjected to SDS-PAGE and Western blot analysis as described previously.11 NIH Image 1.62 was used for densitometric quantification of the bands.
Measurement of NO2/NO3
Confluent HUVECs were treated with M199 plus 10% SCS in the presence and absence of APC for 1 hour at 37°C. Medium was collected and NO2/NO3 levels in culture media were measured using a nitric oxide colorimetric assay kit (Hoffman La Rocsh).
Measurement of cGMP Levels in HUVECs
Confluent HUVECs were treated M199 plus 10% SCS in the presence and absence of APC for 30 minutes at 37°C. After incubation, cells were washed by ice-cold PBS and incubated with ice-cold 65% ethanol. Ethanol solution were collected to new tube, dried under nitrogen, and stored till cGMP measurement. Levels of cGMP were performed according to protocols supplied by the manufacturer (Amersham Pharmacia Biotech).
Mouse Corneal Angiogenesis Assay
The study protocol was approved by the Kumamoto University Animal Care and Use Committee, and care and handing of the animals were in accordence with the National Institutes of Health guidelines. The mouse corneal assay was performed according to procedures described.14 Under sterile conditions, slow-release pellets were prepared incorporating APC or various material into a casting solution of a ethynil-vinyl copolymer (Elvax-40, DuPont), in 10% methylene chloride. Male 8-week-old C57BL/6 mice or eNOS knockout mice (Jackson Laboratory, Bar Harbor, Maine)15 were anesthetized with sodium pentobarbital, corneal micropockets were created in the corneal stroma, and pellets were implanted. The eyes were examined by a slit-lamp biomicroscope (Kowa Company, Ltd., Nagoya, Japan) on day 6 after pellet implantation. Angiogenic responses were evaluated by vessel length and neovascularization area of cornea.16
Statistical analysis was performed with analysis of variance followed by Scheffe’s post hoc test for multiple comparison. A level of P<0.05 was used as statistically significant.
An expanded Materials and Methods section is available in the online data supplement at http://circres.ahajournals.org.
Effects of APC and DIP-APC on MAPK Pathway Activation in Endothelial Cells
To confirm that APC activates the MAPK pathway in endothelial cells, we examined the effect of APC on phosphorylation of MEK1/2 and ERK1/2 in cultured HUVECs in the present study. As shown Figure 1A, intracellular levels of phosphorylated MEK1/2 began to increase 5 minutes after treatment with APC (300 nmol/L), peaking at 15 minutes after treatment, and thereafter gradually decreased to pretreatment levels. Intracellular levels of phosphorylated ERK1/2 began to increase from 5 minutes after treatment with APC, peaking at 15 minutes after the treatment, and thereafter gradually decreased to pretreatment levels (Figure 1A). To determine whether serine protease activity of APC is critical for APC-induced MAPK activation in endothelial cells, we examined the effect of DIP-APC, an active site-blocked APC, on the phosphorylation of ERK1/2 in cultured HUVECs. As shown in Figure 1B, DIP-APC did not increase intracellular levels of phosphorylated ERK1/2.
Effects of APC and DIP-APC on BrdU Incorporation and Proliferation of Endothelial Cells
APC significantly increased BrdU incorporation and proliferation of endothelial cells in a concentration-dependent manner: half-maximal concentrations of APC for BrdU incorporation and proliferation of endothelial cells were 6.5 and 10 nmol/L, respectively. (Figure 1B and 1C). DIP-APC (300 nmol/L) did not increase BrdU incorporation and proliferation of endothelial cells (Figure 1B and 1C), suggesting that serine protease activity of APC might be essential for the induction of DNA synthesis and proliferation of endothelial cells.
Effects of Anti-EPCR Antibodies and an Anti–PAR-1 Antibody on Activation of the MAPK Pathway, DNA Synthesis, and Proliferation Seen in Endothelial Cells Treated With APC
As shown in Figure 2A, anti-EPCR antibody 252, which inhibits the interaction of APC with EPCR,17 significantly reduced APC-induced MAPK activation observed 15 minutes after activation in cultured HUVECs. In contrast, anti-EPCR antibody 92, which does not affect the binding of APC to EPCR,17 did not have any effect (Figure 2A). Increase in BrdU incorporation and proliferation of endothelial cells induced by APC were significantly inhibited by anti-EPCR antibody 252, but not by anti-EPCR antibody 92 (Figure 2B and 2C).
To determine whether PAR-1 plays an important role in APC-induced endothelial MAPK activation and endothelial cell proliferation, we examined the effect of anti-PAR-1 antibody on APC-induced events in cultured HUVECs. Anti-PAR-1 antibody used in the present study completely inhibited endothelial MAPK activation, increase in BrdU incorporation, and proliferation of endothelial cells induced by thrombin (Figure 3). However, although the antibody inhibited endothelial MAPK activation at 5 minutes after treatment with APC (Figure 3A), it did not inhibit MAPK activation at 15 minutes after treatment with APC (Figure 3B). Neither increase in BrdU incorporation nor proliferation of endothelial cells induced by APC was inhibited by anti-PAR-1 antibody (Figure 3C and 3D). These observations suggested that PAR-1 might be implicated in the early activation of MAPK, but might not be critically involved in the late and complete activation of MAPK, leading to DNA synthesis and endothelial proliferation, in cultured HUVECs treated with APC.
Effect of APC on Nitric Oxide Production and Phosphorylation of eNOS in Endothelial Cells
To examine whether APC increases the endothelial nitric oxide (NO) production, we measured NO2/NO3 concentrations in culture media of HUVECs treated with APC. Levels of NO2/NO3 in culture media of HUVECs at 1 hour after addition of APC (300 nmol/L) (10.0±1.9 μmol/L) were significantly higher than those at 1 hour after addition of saline (6.7±1.3 μmol/L, P<0.01).
To determine whether APC might activate eNOS, thereby increasing NO production in endothelial cells, we examined the effect of APC on phosphorylated eNOS levels in endothelial cells. As shown in Figure 4A, APC increased phosphorylation of eNOS. APC-induced phosphorylation of eNOS was inhibited by LY-294,002 and wortmannin, both of which are inhibitors of PI3-kinase (Figure 4B).
Effects of l-NAME, LY-294,002, and Wortmannin on MAPK Activation, Increase in BrdU Incorporation, and Proliferation Seen in Endothelial Cells Treated With APC
To determine whether NO plays a role in APC-induced MAPK activation and endothelial cell proliferation, we examined the effect of l-NAME, an inhibitor of NOS, on endothelial MAPK activation, DNA synthesis, and proliferation seen in endothelial cells treated with APC. As shown in Figure 5, l-NAME inhibited these changes induced by APC. Furthermore, LY-294,002 and wortmannin also inhibited MAPK activation, increase in BrdU incorporation, and proliferation seen in endothelial cells treated with APC.
These observations strongly suggested that APC might activate eNOS via PI3-kinase–dependent phosphorylation, thereby increasing NO production, and the NO thus formed might play a critical role in the endothelial MAPK activation, DNA synthesis, and proliferation seen in endothelial cells treated with APC.
Effect of APC on Intracellular Levels of cGMP and Effect of KT5823, a PKG Inhibitor, on the Endothelial MAPK Activation, Increase in BrdU Incorporation, and Proliferation Seen in Endothelial Cells Treated With APC
Intracellular levels of cGMP in endothelial cells were increased by APC treatment (1.78±0.65 fmol/well in control versus 3.13±0.83 fmol/well in APC treatment; P<0.01). KT5823, an inhibitor of PKG, inhibited MAPK activation, increase in BrdU incorporation and proliferation seen in endothelial cells treated with APC (Figure 5). These observations strongly suggested that APC might activate the MAPK pathway, thereby inducing DNA synthesis, endothelial cell proliferation, and MAPK activation could be induced by PKG, which was activated via an NO-dependent increase in intracellular levels of cGMP in endothelial cells treated with APC.
APC Induces Morphological Differentiation
To examine whether APC induces morphogenetic changes resembling tube-like structures, we investigated the effect of APC on the morphological change of HUVECs plated on Matrigel. Treatment with APC for 16 hours resulted in morphological changes, including elongation and formation of thin cords of interconnecting cells (Figure 6B and 6D). These changes were not induced by DIP-APC treatment (Figure 6C and 6D).
Effect of APC on Angiogenesis In Vivo
Because APC induced the endothelial cell proliferation by activating the MAPK pathway in cultured HUVECs as shown in the present study, it is possible that APC induces angiogenesis in vivo. We examined this possibility using the mouse corneal angiogenesis model. As shown in Figure 7C, APC induced corneal angiogenic response; ie, new microvessels crossed the cornea from the limbus toward the pellet containing APC. Angiogenic response was not observed with the pellet alone (Figure 7A). In this experimental condition, APC-induced corneal angiogenic response was comparable to that of VEGF (Figure 7B). DIP-PAC, active site blocked APC, did not show any angiogenic response (Figure 7D). APC-induced corneal angiogenic response was not observed in eNOS knockout mice (Figure 7F).
In the present study, we demonstrated that APC induced activation of both MEK1/2 and ERK1/2, increase in DNA synthesis, and proliferation in cultured HUVECs. Because activation of this MAPK system has been shown to be a prerequisite for growth factor–stimulated mitogenesis in many cells,12 endothelial cell proliferation induced by APC might be at least partly explained by activation of these enzymes.
Phosphorylation of eNOS was increased by APC, which was inhibited by LY-294,002 and wortmannin. Furthermore, l-NAME, LY-294,002, and wortmannin inhibited both MAPK activation and endothelial cell proliferation induced by APC. Nitric oxide emerged as an important upstream signal of cell proliferation in endothelial cells.18 Hood et al12 reported that nitric oxide plays critical roles in the signaling pathways and endothelial cell proliferation induced by VEGF. These observations strongly suggested that APC might activate eNOS via PI3-kinase–dependent phosphorylation, thereby inducing both MAPK activation and endothelial cell proliferation.
The precise mechanism(s) by which APC signals in cultured HUVECs, thereby activating the MAPK pathway, is not fully elucidated. Recently, Riewald et al6 reported that APC activates PAR-1 by interacting with EPCR on endothelial cells. They also showed that EPCR-bound APC induced phosphorylation of MAPK via PAR-1 activation. These observations are consistent with our present observations that endothelial MAPK activation was inhibited by anti-EPCR antibody 252 that is capable of inhibiting the interaction of APC with EPCR and that the activation was not induced by DIP-APC, an inactive derivative of APC. Riewald et al6 also demonstrated that anti–PAR-1 antibody significantly inhibited MAPK activation induced by APC observed at 6 minutes after treatment with APC. Consistent with this observation, anti–PAR-1 antibody significantly inhibited APC-induced MAPK activation in cultured HUVECs 5 minutes after treatment with APC as shown in the present study. However, anti–PAR-1 antibody inhibited neither MAPK activation at 15 minutes after treatment of endothelial cells with APC, increase in DNA synthesis, nor proliferation in cultured HUVECs treated with APC. Because this antibody completely inhibited both MAPK activation and endothelial cell proliferation induced by thrombin as shown in the present study, PAR-1–mediated MAPK activation and the subsequent endothelial cell proliferation could be completely inhibited by the antibody in the present experimental conditions. These observations suggested that PAR-1 might be implicated in early MAPK activation, but might not be critically involved in the late and complete activation of MAPK leading to endothelial proliferation in cultured HUVECs treated with APC. Further studies are necessary to determine the mechanism(s) by which APC signals the endothelial cells to activate the MAPK pathway.
Recently, Domotor et al19 reported that EPCR-bound APC induced intracellular calcium mobilization from the endoplasmic reticulum in cultured human umbilical vein endothelial cells. Because ionomycin-induced calcium release activates the ERK1/2 in cultured HUVECs20 and thrombin-induced MAPK activation was inhibited by the intracellular calcium chelator, 1,2-bis(o-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid tetra (acetoxymethyl) ester (BAPTA-AM),20 it is possible that APC activates the MAPK pathway by inducing intracellular calcium mobilization. However, this possibility seems less likely because APC-induced intracellular calcium mobilization was shown to be a PAR-1–dependent process.19 Furthermore, observations in our preliminary experiments demonstrated that APC-induced MAPK activation was only partially inhibited by BAPTA-AM, whereas thrombin-induced MAPK activation was almost completely inhibited by BAPTA-AM, suggesting that APC-induced MAPK activation might not be mediated by calcium mobilization in cultured HUVECs in the present study.
In normal human subjects, plasma levels of protein C and APC are reported to be 65 nmol/L and 38 pmol/L, respectively.21 Because APC is formed by thrombin bound to thrombomodulin on the endothelial surface,1 local concentration of APC might be increased at the site where thrombin generation is increased. Consistent with this hypothesis, Hanson et al22 reported that plasma levels of APC were found to be 8 nmol/L in baboons given thrombin intravenously. Because 8 nmol/L was a half-maximal concentration of APC required for the induction of increase in DNA synthesis and endothelial cell proliferation as shown in the present study, it is possible that APC induces endothelial cell proliferation leading to angiogenesis at the site where thrombin generation is increased.
Physiological relevance of angiogenesis induced by APC is not clear at present. Development of the vascular system in various organs in protein C knockout mice was reported to be normal during embryogenesis.23 EPCR knockout mice showed early embryonic lethality, but normal vascular system development.24 These results suggested that APC might not be involved in vasculogenesis and embryonic angiogenesis.
Postneonatal angiogenesis, referred to as neovascularization, was observed in various physiological and pathological conditions such as wound healing, postischemic tissue restoration, and carcinogenesis.25 Expression of EPCR was demonstrated to be enhanced in carcinomatous tissues of breast cancer, melanomas, and renal and colon carcinomas,26,27 suggesting that APC might be critically involved in neovascularization in these carcinomatous tissues. Neovascularization is also observed in ischemic tissues.28 Hypoxia activates the coagulation cascade by increasing the endothelial expression of tissue factor, a triggering substance for the coagulation cascade, leading to thrombin generation.29 Thrombin thus generated binds to thrombomodulin, thereby activating protein C to form APC,3 suggesting that generation of APC might be increased in the ischemic tissue. APC was shown to be beneficial in reducing ischemia/reperfusion-induced injury of the kidney,30 spinal cord,9,31 and brain17 based on its antiinflammatory properties. In addition, such therapeutic effects of APC might be at least partly explained by angiogenic activity, leading to the neovascularization by which ischemic tissue injury can be attenuated through restoration of the blood supply. This possibility should be further investigated using protein C knockout mice subjected to the ischemia-induced tissue injury in the near future.
This study was supported in part by research grants from the departmental funds of Graduate School of Medical Sciences, Kumamoto University.
Original received December 19, 2003; revision received May 14, 2004; accepted May 18, 2004.
Esmon CT. Thrombomodulin as a model of molecular mechanisms that modulate protease specificity and function at the vessel surface. FASEB J. 1995; 9: 946–955.
Stearns-Kurosawa DJ, Kurosawa S, Mollica JS, Ferrell GL, Esmon, CT. The endothelial cell protein C receptor augments protein C activation by the thrombin-thrombomodulin complex. Proc Natl Acad Sci U S A. 1996; 93: 10212–10216.
Esmon CT. The protein C anticoagulant pathway. Arterioscler Thromb. 1992; 12: 135–145.
Sakata Y, Loskutoff DJ, Gladson CL, Hekman CM, Griffin JH. Mechanism of protein C-dependent clot lysis: role of plasminogen activator inhibitor. Blood. 1986; 68: 1218–1223.
Riewald M, Petrovan RJ, Donner A, Mueller BM, Ruf W. Activation of endothelial cell protease activated receptor 1 by the protein C pathway. Science. 2002; 296: 1880–1882.
Pages G, Lenormand P, L’Allemain G, Chambard JC, Meloche S, Pouyssegur J. Mitogen-activated protein kinases p42mapk and p44mapk are required for fibroblast proliferation. Proc Natl Acad Sci U S A. 1993; 90: 8319–8323.
Hood J, Granger HJ. Protein kinase G mediates vascular endothelial growth factor-induced Raf-1 activation and proliferation in human endothelial cells. J Biol Chem. 1998; 273: 23504–23508.
Meyer M, Clauss M, Lepple-Wienhues A, Waltenberger J, Augustin HG, Ziche M, Lanz C, Buttner M, Rziha HJ, Dehio C. A novel vascular endothelial growth factor encoded by Orf virus, VEGF-E, mediates angiogenesis via signalling through VEGFR-2 (KDR) but not VEGFR-1 (Flt-1) receptor tyrosine kinases. EMBO J. 1999; 18: 363–374.
Cao R, Brakenhielm E, Wahlestedt C, Thyberg J, Cao Y. Leptin induces vascular permeability and synergistically stimulates angiogenesis with FGF-2 and VEGF. Proc Natl Acad Sci U S A. 2001; 98: 6390–6395.
Parenti A, Morbidelli L, Cui XL, Douglas JG, Hood JD, Granger HJ, Ledda F, Ziche M. Nitric oxide is an upstream signal of vascular endothelial growth factor-induced extracellular signal-regulated kinase 1/2 activation in postcapillary endothelium. J Biol Chem. 1998; 273: 4220–4226.
Domotor E, Benzakour O, Griffin JH, Yule D, Fukudome K, Zlokovic BV. Activated protein C alters cytosolic calcium flux in human brain endothelium via binding to endothelial protein C receptor and activation of protease activated receptor-1. Blood. 2003; 101: 4797–4801.
Gruber A, Griffin JH. Direct detection of activated protein C in blood from human subjects. Blood. 1992; 79: 2340–2348.
Gu JM, Crawley JT, Ferrell G, Zhang F, Li W, Esmon NL, Esmon CT. Disruption of the endothelial cell protein C receptor gene in mice causes placental thrombosis and early embryonic lethality. J Biol Chem. 2002; 277: 43335–43343.
Felmeden DC, Blann AD, Lip GY. Angiogenesis: basic pathophysiology and implications for disease. Eur Heart J. 2003; 24: 586–603.
Mizutani A, Okajima K, Uchiba M, Noguchi T. Activated protein C reduces ischemia/reperfusion-induced renal injury in rats by inhibiting leukocyte activation. Blood. 2000; 95: 3781–3787.
Taoka Y, Okajima K, Uchiba M, Murakami K, Harada N, Johno M, Naruo M. Activated protein C reduces the severity of compression-induced spinal cord injury in rats by inhibiting activation of leukocytes. J Neurosci. 1998; 18: 1393–1398.